What Is Anthropic Mythos? The AI Cybersecurity Model That Has Concerned FM Nirmala Sitharaman

Union Finance Minister Nirmala Sitharaman has flagged concerns over Anthropic’s AI system Mythos, describing the emerging cyber risk linked to it as a threat comparable to war. Her remarks have brought attention to a relatively lesser-known but highly advanced cybersecurity-focused AI model.

“No one would have imagined a couple of weeks ago that there is a new threat which is as big as a threat of war, that is going to hit us in terms of the entire digital network,” said Sitharaman, while speaking at the ET Awards for Corporate Excellence.

So, what exactly is Mythos?

Mythos is an unreleased preview AI model developed by Anthropic as part of Project Glasswing. Unlike general-purpose chatbots designed for writing, coding or research, Mythos is reportedly built specifically

for defensive cybersecurity operations. It is being made available only to a limited set of organisations under controlled testing conditions.

The model is understood to help detect vulnerabilities, simulate cyberattacks, analyse threat patterns and strengthen digital defence systems at machine speed. In simple terms, it is an AI tool designed to protect networks before hackers can exploit weaknesses.

Why Has It Triggered Concern?

While such systems can be powerful for defence, governments worldwide are also wary of how advanced cybersecurity AI could be misused if it falls into the wrong hands. A model capable of identifying weaknesses in networks could potentially be weaponised to launch sophisticated attacks, disrupt digital infrastructure or overwhelm critical systems.

Anthropic in its blogpost said, “During our testing, we found that Mythos Preview is capable of identifying and then exploiting zero-day vulnerabilities in every major operating system and every major web browser when directed by a user to do so. The vulnerabilities it finds are often subtle or difficult to detect. Many of them are ten or twenty years old, with the oldest we have found so far being a now-patched 27-year-old bug in OpenBSD—an operating system known primarily for its security.”

According to the blogpost, Anthropic Mythos Preview found a 27-year-old bug in OpenBSD that nobody had noticed ever, found a 16-year-old bug in FFmpeg (used by almost every video app on the planet), found bugs in every major browser and every major operating system.

It not just found these vulnerabilities but also walked through it. In layman terms, it didn’t just find the unlocked door — it figured out how to walk through it and take over the building.

“Over 99% of the vulnerabilities we’ve found have not yet been patched,” it said.

It also said non-experts can also leverage Mythos Preview to find and exploit sophisticated vulnerabilities.

Previously only elite hackers with years of experience could do this. Now, someone with no security training can ask the AI and wake up to a working cyberattack.

That appears to be the basis of Sitharaman’s warning. India has rapidly digitised banking, payments, governance services and business systems in recent years. This creates enormous benefits, but it also means cyber threats can now have economy-wide consequences.

“We will just have to keep exceptionally vigilant,” she said.

India’s Response

According to Sitharaman, the government has already stepped up engagement with Anthropic and the US administration to better understand the technology and its implications. The Ministry of Electronics and Information Technology is reviewing the issue, while officials have begun reaching out to organisations that received access to Mythos.

The Centre has also consulted banks and financial institutions, reflecting concern over risks to the financial system. A specialised panel, headed by C S Setty of State Bank of India, has been announced after a high-level meeting in New Delhi attended by Sitharaman and IT Minister Ashwini Vaishnaw.

Why This Matters

Mythos may still be in preview mode, but the debate around it signals a bigger global shift: AI is no longer just about productivity and chatbots. It is increasingly becoming central to cybersecurity, national resilience and digital warfare preparedness.

For India, Sitharaman’s remarks underline that protecting digital infrastructure may now be as important as protecting physical borders.